Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Grok 3 [Beta] is clearly ahead on the aggregate, 38 to 33. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Grok 3 [Beta]'s sharpest advantage is in multimodal & grounded, where it averages 43.2 against 32.4. The single biggest benchmark swing on the page is HumanEval, 34 to 17. LFM2.5-1.2B-Thinking does hit back in instruction following, so the answer changes if that is the part of the workload you care about most.
LFM2.5-1.2B-Thinking is the reasoning model in the pair, while Grok 3 [Beta] is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use. Grok 3 [Beta] gives you the larger context window at 128K, compared with 32K for LFM2.5-1.2B-Thinking.
Pick Grok 3 [Beta] if you want the stronger benchmark profile. LFM2.5-1.2B-Thinking only becomes the better choice if instruction following is the priority or you want the stronger reasoning-first profile.
Grok 3 [Beta]
35.7
LFM2.5-1.2B-Thinking
34.1
Grok 3 [Beta]
16.3
LFM2.5-1.2B-Thinking
8.2
Grok 3 [Beta]
43.2
LFM2.5-1.2B-Thinking
32.4
Grok 3 [Beta]
47.4
LFM2.5-1.2B-Thinking
38.4
Grok 3 [Beta]
34
LFM2.5-1.2B-Thinking
27
Grok 3 [Beta]
67
LFM2.5-1.2B-Thinking
72
Grok 3 [Beta]
58.7
LFM2.5-1.2B-Thinking
60.7
Grok 3 [Beta]
49.1
LFM2.5-1.2B-Thinking
42.3
Grok 3 [Beta] is ahead overall, 38 to 33. The biggest single separator in this matchup is HumanEval, where the scores are 34 and 17.
Grok 3 [Beta] has the edge for knowledge tasks in this comparison, averaging 34 versus 27. Inside this category, MMLU is the benchmark that creates the most daylight between them.
Grok 3 [Beta] has the edge for coding in this comparison, averaging 16.3 versus 8.2. Inside this category, HumanEval is the benchmark that creates the most daylight between them.
Grok 3 [Beta] has the edge for math in this comparison, averaging 49.1 versus 42.3. Inside this category, AIME 2023 is the benchmark that creates the most daylight between them.
Grok 3 [Beta] has the edge for reasoning in this comparison, averaging 47.4 versus 38.4. Inside this category, LongBench v2 is the benchmark that creates the most daylight between them.
Grok 3 [Beta] has the edge for agentic tasks in this comparison, averaging 35.7 versus 34.1. Inside this category, BrowseComp is the benchmark that creates the most daylight between them.
Grok 3 [Beta] has the edge for multimodal and grounded tasks in this comparison, averaging 43.2 versus 32.4. Inside this category, MMMU-Pro is the benchmark that creates the most daylight between them.
LFM2.5-1.2B-Thinking has the edge for instruction following in this comparison, averaging 72 versus 67. Inside this category, IFEval is the benchmark that creates the most daylight between them.
LFM2.5-1.2B-Thinking has the edge for multilingual tasks in this comparison, averaging 60.7 versus 58.7. Inside this category, MGSM is the benchmark that creates the most daylight between them.
Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.
Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.